Video monitoring devices allow individuals and businesses to monitor premises for various purposes, including, for example, security, baby or elderly monitoring, videoconference, etc. Such video monitoring devices may record videos continuously, generating a huge amount of video data every day. Reviewing video data, however, may be challenging. For example, a user may not have enough time to review a video in its entirety.
Such inconvenience may be partially resolved by displaying some video preview frames extracted from the video so that a user can review the video preview frames instead of the whole video. For example, a video preview frame may be selected from the video every certain period of time. Although this may be easy to implement, there are shortcomings. The extracted video preview frames may not catch all special events (e.g., a baby crying). Thus, a user who only reviews these video preview frames may miss some special events. In addition, the video preview frames presented to the user may look the same, and the user may still miss a special event included in the video preview frames if there is no indication that the special event occurred.